Fifty four consecutive patients with oesophageal variceal bleeding were randomised to undergo intravariceal (28 patients) or paravariceal (26 patients) sclerotherapy, every three weeks. Intravariceal technique was found significantly (p less than 0.01) more effective in controlling active variceal bleeding than the paravariceal technique (91% v 18.7% respectively). The mean (+/- SD) time taken for variceal eradication by intravariceal sclerotherapy (15.4 +/- 5.3 weeks) was significantly (p less than 0.001) less than paravariceal (26.8 +/- 6.6 weeks) technique. The number of sclerotherapy sessions needed with intravariceal technique were also significantly less. Rebleeding was seen in 38.5% patients after para and 14.3% after intravariceal injections (NS). Except for retrosternal pain, which occurred more often (p less than 0.01) with paravariceal technique, there was no difference in the incidence of other complications or mortality between the two groups. Variceal recurrence was seen in seven patients (25%) in the intra and one (3.9%) patient in the paravariceal group (p less than 0.01) after a mean follow up of 29.4 +/- 9.1 weeks. Intravariceal sclerotherapy was superior to paravariceal in the control of active variceal bleeding and for total variceal obliteration, but was associated with a higher variceal recurrence.